A Comparative study of Digital Watermarking algorithms DWT ... · (DCT), Discrete Wavelet Transform...

A Comparative study of Digital Watermarking algorithms DWT, SVD &

DWT-SVD in Medical Field

Jaee P. Gaikwad

PG Student, Department of Electronics Engineering,

D.Y.Patil College of Engg. & Technology,Kasaba Bawada Kolhapur,

Maharashtra, India.

Dr.Mrs.K.V.Kulhalli Vice Principal and HOD, Department of Information & Technology,

D.Y.Patil College of Engg. & Technology,Kasaba Bawada Kolhapur,

Maharashtra, India.

Prof. S.R.Khot

Associate Professor, Department of Electronics Engineering,

D.Y.Patil College of Engg. & Technology,Kasaba Bawada Kolhapur, Maharashtra, India.

Abstract Information and Communication Technologies (ICT) are being adopted widely to improve citizen’s health care. Health

information systems (HIS) of different hospitals exchange

electronic medical records including digital medical images of

patients. Medical images and accompanying reports have

special requirement of protecting the privacy of the patient by

not revealing personal particulars especially when they are

transmitted over networks. To increase the security of medical

images and preserve patients’ privacy, digital watermarking

has been proposed. The aim of digital watermarking is to hide

some secret information or logo into the multimedia content

for protecting the content from unauthorized access or illegal use. Digital image watermarking is a promising domain for

various applications, for example, ownership identification,

copy protection, authentication, broadcast monitoring, tamper

detection & recovery etc. In this paper we are going to

compare three different techniques used in digital

watermarking. They are DWT,SVD & hybrid DWT-SVD[2].

To evaluate their performance, these schemes are

exposed to different geometric and non-geometric attacks.

The comparison is made in terms of their performance to

sustain to attack. To check effectiveness of these techniques

for imperceptibility and robustness, PSNR and NCC parameters are used. The quality of the imperceptibility of the

system is calculated by the Peak Signal to Noise Ratio of the

watermarked image with original image. The similarity

between inserted and extracted watermark is estimated by

Normalized Correlation Coefficient.

Keywords: Medical Image watermarking, Discrete Wavelet

Transform,Singular Value Decomposition,Peak Signal to

NoiseRatio,Normalized(PSNR),Correlation Coefficient(NCC)

Introduction Exchange of medical images between hospitals has become a natural practice of modern times. The medical images are

exchanged for a variety of reasons like teleconferences among

clinicians, to discuss diagnostic and therapeutic measures and

so on. This exchange of medical images inflicts three

restraints for the medical images: (1)only authorized persons

have right to use the information,(2) the information has not

been changed by unauthorized users and (3) there should be

evidence that the information belongs to the correct patient [].

On the other hand transmission of medical image and patient

data separately through commercial networks like internet

results in excessive cost and transmission time. Watermarking is one of the techniques used to address the above two

issues[3].

According to the domain in which the watermark is

inserted, these techniques are classified into two categories,

i.e., spatial domain and transform domain methods. The

spatial domain methods modify the digital data (pixels)

directly to hide the watermark bits and possess the advantage

of low computational complexity. On the other hand, the

transform (frequency) domain methods do not alter the pixel

values directly but rather modify the transform coefficients to

hide the watermark bits such as Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT) and Singular

Value Decomposition (SVD)[2]. The rest of the paper is organized as follows. Section

II provides brief details about the DWT, SVD and DWT-SVD

based watermarking algorithms. Experimental study and

results are given in Section III. Section IV gives the

conclusion & future work.

International Journal of Engineering Research and Technology. ISSN 0974-3154 Volume 10, Number 1 (2017) © International Research Publication House http://www.irphouse.com

859

Methodology Discrete Wavelet Transform (DWT):This is a frequency

domain technique in which firstly cover image is transformed

into frequency domain and then its frequency coefficients are modified in accordance with the transformed coefficients of

the watermark and watermarked image is obtained which is

very much robust. In single level decomposition, DWT

decomposes image hierarchically, providing both spatial and

frequency description of the image. It decompose an image in

basically three spatial directions i.e., horizontal, vertical and

diagonal in result separating the image into four different

components namely LL, LH, HL and HH. Here first letter

refers to applying either low pass frequency operation or high

pass frequency operations to the rows and the second letter

refers to the filter applied to the columns of the cover image. LL level is the lowest resolution level which consists of the

approximation part of the cover image. Rest three levels i.e.,

LH, HL, HH give the detailed information of the cover image.

DWT Embedding Algorithm: The embedding algorithm for

DWT based watermarking is shown in Figure 1. The

algorithm works as follows:

Step 1: The original N*N RGB image is transformed into sub-

bands using single level 2-D DWT.

Step 2: The watermark of size M*M RGB image is

transformed into sub-bands using single level 2-D DWT.

Step 3: The resultant watermark is then embedded into the lower frequency sub-band of original image using the scale

factor (α) i.e. WI=O+ αW

Step 4: Finally, inverse 2-D DWT is performed to produce the

watermarked image[4].

Fig 1: DWT based Embedding

DWT Extraction Algorithm: The extraction algorithm for

DWT based watermarking is shown in Figure 2. The

algorithm works as follows:




transformed into sub-bands using single level 2-D DWT.

Step 3: The watermarked image (output of embedding) is

transformed into sub-bands using the single level 2-D DWT.

Step 4: Then the extraction is applied to the decomposed watermarked image using the same value of scale factor (α)

i.e. EWI=(WM – O)/ α

Step 5: Finally, inverse 2-D DWT is performed to get the

extracted watermark image.

Fig 2: DWT based Extraction

Singular Value Decomposition (SVD): Singular Value

Decomposition is a linear algebra transform which is used for

factorization of a real or complex matrix with numerous

applications in various fields of image processing. As a digital

image can be represented in a matrix form with its entries

giving the intensity value of each pixel in the image, SVD of

an image M with dimensions m x m is given by : M=USVT

Where, U and V are orthogonal matrices and S known as singular matrix is a diagonal matrix carrying non-negative

singular values of matrix M. The columns of U and V are

called left and right singular vectors of M, respectively. They

basically specify the geometry details of the original image.

Left singular matrix i.e., U represents the horizontal details

and right singular matrix i.e., V represents the vertical details

of the original image. The diagonal values of matrix S are

arranged in decreasing order which signifies that the

importance of the entries is decreasing from first singular

value to the last one. This feature is employed in SVD based

compression techniques[1]. There are two main properties of SVD to employ in digital

watermarking schemes:

1. Small variations in singular values do not affect the quality

of image.

2. Singular values of an image have high stability.

Hybrid DWT-SVD: Hybrid technique is a fusion of two

techniques. Here, DWT and SVD are used together to

improve the quality of digital watermarking and hence increases the robustness and imperceptibility of an image.

Hybrid DWT-SVD Embedding Algorithm: The embedding

algorithm for DWT-SVD based watermarking is shown in

Figure 3. The algorithm works as follows:



Step 2: SVD is performed on LL sub-band (on RGB

components) of decomposed RGB original image i.e.,

S=USVT


transformed into sub-bands using single level 2-D DWT. Step 4: SVD is performed on LL sub-band (on RGB

components) of decomposed RGB watermark image i.e., SW=

UwSwVwT


860

Step 5: After performing SVD on both original and watermark

images, the resultant watermark image is then embedded with

the original image using the scale factor (α) i.e., SWI=S + α

(SW)

Step 6: Inverse SVD is performed on embedded image.

Step 7: Finally, inverse 2-D DWT is performed to produce the watermarked image.

Fig 3: DWT-SVD based Embedding

Hybrid DWT-SVD Watermark Extraction Algorithm: The

extraction algorithm for DWT-SVD based watermarking is

shown in Figure 4. The algorithm works as follows: Step 1: The original N*N RGB image is transformed into sub-bands using single level 2-D DWT. Step 2: SVD is performed on LL sub-band (on RGB components) of decomposed RGB original image i.e., S = USVT Step 3: The watermark of size M*M RGB image is transformed into sub-bands using single level 2-D DWT. Step 4: SVD is performed on LL sub-band (on RGB components) of

decomposed RGB watermark image i.e., SW = UwSwVwT

Step 5: The watermarked image (output of embedding) is transformed into sub-bands using the single level 2-D DWT. Step 6: SVD is performed on LL sub-band (on RGB components) of decomposed RGB watermarked image i.e., SWI = UwSwVw

T Step 7: Then the extraction is applied to the resultant SVD image using the same value of scale factor (α) i.e., EWI= (SWI – S) / α Step 8: Inverse SVD is applied on resultant image after extraction.

Step 9: Finally, inverse 2-D DWT is performed to get the extracted watermark image.

Fig 4: DWT-SVD based Extraction

Experimental Results For experimental study, we have used ultrasonic medical

images as host image in which watermark is embedded.

ROI(Region of Interest) & EPR(Electronic Patient Record)

are concatenated & used as watermark.

Sample 1 Sample 2 Sample 3 Sample 4

Fig 5: Sample Medical Images used as Host Image

Watermark 1 Watermark 2

Fig 6: Watermark Images


861

Fig 7:GUI for Embedding & Extraction of watermark

Table 1: Comparison of PSNR & NCC values

Fig 8:Graph of sample Images against PSNR values

Fig 9:Graph of sample Images against NCC values

Fig 10: GUI for applying attacks on watermarked image &

extracted watermark from attacked watermarked image

Medical

Images

(Host

Image)

Concatenated Watermark

DWT SVD DWT-SVD

PSNR NCC PSNR NCC PSNR NCC

Sample 1 55.29 0.990 67.710 0.8864 35.294 0.9901

Sample 2 54.21 0.985 67.492 0.8650 34.214 0.9859

Sample 3 55.7 0.988 70.190 0.8945 35.7 0.9889

Sample 4 54.91 0.957 70.114 0.8871 34.910 0.9576

Sample 5 55.40 0.989 70.921 0.8878 35.408 0.9894


862

Type of

Attack

DWT SVD DWT-SVD

Mean

Attacked

Image

Extracted watermark

Salt &

Pepper

Attacked

Image

Extracted watermark

Rotation

Attacked

Image

Extracted watermark

Blurring

Attacked

Image

Extracted watermark

Fig 11: Comparison by various attacks

Fig 12:Graph of Type of attacks against PSNR values

Fig 13:Graph of Type of attacks against NCC values


863

Table 2:Comparison of PSNR & NCC values against various

attacks

In the evaluation of the performance of the watermarking

schemes, we calculate the peak signal to noise ratio (PSNR)

between the original and watermarked images and

.normalized correlation coefficient between the original and

extracted watermark.Results are also compared by applying

various attacks on watermarked images.

Conclusion and Future Work Medical images may carry sensitive information and

additional security measures are required to preserve the

privacy of patients. Invisible watermarking has emerged as an

effective technique to achieve this goal. In this paper, we have

implemented and studied three different and popular

watermarking algorithms viz. DWT, SVD and DWT-SVD

based watermarking algorithms.We have embedded two

different watermarks to test the capacity of the algorithms. We

have assessed the quality of the watermarked images using

PSNR and NCC. From objective analysis, we found that watermarking has not resulted in any loss of medical

information and also, watermarked images are similar to the

original images. Further, we conducted different attacks to

check the robustness of the watermarking algorithms.

In future work, instead of embedding watermark

directly, sparse coding will be applied on concatenated

watermark image.Using sparse code,size of watermark is

reduced hence only small bits of information is embedded.so

quality of watermarked image & extracted watermark will be

enhanced.Also if watermarked image undergo various attacks

& get tampered then image will be recovered without loss of

any information.

References [1]Afaf Tareef, Ahmad Al-Ani,Hung Nguyen, Yuk Ying

Chung,“A Novel Tamper Detection-Recovery and

Watermarking System for Medical Image Authentication and

EPR Hiding”, IEEE 2014.

[2]Sachin Mehta, RajarathnamNallusamy, Ranjeet Vinayak

Marawar, Balakrishnan Prabhakar “A study of DWT and SVD

basedWatermarking Algorithms for Patient Privacy in

Medical Images”, IEEE International Conference on

Healthcare Informatics, 2013.

[3]R. Eswaraiah, E. Sreenivasa Reddy, “ A Fragile ROI-Based

Medical Image Watermarking Technique with Tamper Detectionband Recovery” 2014 Fourth International

Conference on Communication Systems and Network

Technologies

[4]Nidhi Bisla, Prachi Chaudhary,“Comparative Study of

DWT and DWT-SVD Image Watermarking Techniques”

ijarcsse, Volume 3, Issue 6, June 2013

[5]Abhilasha Singh, Malay Kishore Dutta,“ A Blind & Fragile

Watermarking Scheme for Tamper Detection of Medical

Images Preserving ROI”, 2014 International Conference on

Medical Imaging, m-Health and Emerging Communication

Systems (MedCom),2014.

[6]Hui Liang Khor ,Siau-ChuinLiew and JasniMohd. Zain, “A Review of Reversible Medical Image Watermarking Scheme

with Tamper Localization and Recovery Capability”, IEEE

2014 International Conference on Computer, Communication,

and Control Technology (I4CT 2014), September 2 -4, 2014 -

Langkawi, Kedah, Malaysia.

[7]Malay Kishore Dutta, Anushikha Singh &M.Parthasarathi,

Carlos M. Travieso, “Imperceptible Digital Watermarking in

Medical Retinal Images for Tele-Medicine

Applications”,2014.

[8]Yatindrapathak,Satishdehariya, “A More Secure

transmission of medical images by Two Label DWT and SVD based watermarking technique”IEEE International Conference

on Advances in Engineering & Technology Research

(ICAETR - 2014), August 01-02, 2014, Dr. Virendra Swarup

Group of Institutions, Unnao, India.

[9] Praveen Kumar E, Remya Elizabeth Philip, Sunil Kumar

P, Sumithra M G, “DWT-SVD based reversible watermarking

algorithm for embedding the secret data in medical images”,

4th ICCCNT 2013 July 4-6, 2013, Tiruchengode, India.

Type of

Attack

DWT SVD DWT-SVD

PSNR NCC PSNR NCC PSNR NCC

Mean 33.860 0.1284 33.953 0.0895 30.046 0.7310

Salt &

pepper

21.885 0.2758 17.044 0.1702 21.751 0.6095

Rotation 17.079 0.3519 12.306 0.0568 16.587 0.6097

Blurring 31.621 0.5181 26.871 0.0245 30.990 0.5971


864

A Comparative study of Digital Watermarking algorithms DWT ... · (DCT), Discrete Wavelet Transform...

Documents

Transcript of A Comparative study of Digital Watermarking algorithms DWT ... · (DCT), Discrete Wavelet Transform...